The proposed research is designed to investigate the onset and development of those perceptual competencies that must be implemented when the three-dimensional form of objects and the motion of the self are extracted from continuously changing visual information. These two accomplishments, perceiving form and self motion, are assumed to be fundamental requisites of an adaptively functional visual system. The principle objective is to use dynamic processing formalisms, that have recently been forwarded in the literatures on human perception and computer vision, as guides to assessing the infant's developing visual competencies. The research plan includes 21 experiments divided among three related issues. The first series of experiments is designed to examine the onset and early development of those basic processing assumptions needed to extract coherent, three-dimensional forms from the optical flow found in dynamic environments. The second group of experiments seeks to examine how the infant's general and specific world knowledge affects the perceived functional significance of the information extracted from optic flow. In particular, these experiments are concerned with the recognition of point-light displays depicting people walking and faces expressing discrete emotions. The last group of experiments are concerned with the perception of self motion from continuous perspective transformations of the optic array. These experiments will explore the functional significance of this dynamic information for infants' maintenance of postural stability. The studies are designed to investigate the formal properties in dynamic flowfields that are detected by the infant, and also the special role that may be played by peripherally presented information. These experiments will focus on perceptual development during the first year of life. A variety of methods and measures will be used - infant-control habituation paradigm, paired preferential looking techniques, instrumental conditioning, visual fixation, heart rate, postural stability - in order to make use of the most sensitive assessments possible. The results from these experiments will contribute to our understanding of how the infant is initially prepared to extract some of the most general regularities manifest in continuously changing visual information and how experience leads to the development of additional information extraction processes.